Purpose: compare linear and curvilinear models to describe the force-velocity relationship, especially on the velocity end, in lower limb acyclic extensions. Methods: nine athletes performed horizontal lower limb extensions on a leg press ergometer allowing to set from very high to very low (i.e. horizontal assisted extensions) resistive and inertial conditions. Lower limb force and velocity were continuously measured over the push-off in six resistive conditions to determine individual force-velocity relationships. Goodness of fit (GoF) of the linear model on force and velocity data (using the basic first order polynomial function), was compared to that of curvilinear models (using the basic second order polynomial function, Fenn and Marsh’s, and Hill’s equations) based on the Akaike Information Criterion (AIC). Results: When expressed relative to the theoretical maximal force (F0-L) and velocity (v0) obtained from the linear model, force and velocity data ranged from 26.6±6.6 to 96.0±3.6%F0-L and from 8.3±1.9 to 76.6±7.0%v0-L (range of individual values=5-86%v0-L and 16-99%F0-L). Curvilinear and linear models showed very high GoF (r2=0.951-0.999; SEE=17-159N). Despite higher GoF for curvilinear models, AIC showed that the linear model presented ~99-100 % chance to be the best one to describe the force-velocity relationship. Conclusion: the force-velocity relationship in acyclic lower limb extensions can be considered as linear from its force (from 5%v0-L) to its velocity end (up to 86%v0-L). Assisted horizontal acyclic lower limb extensions allowed achievement of movement velocities very close to v0, representing an interesting modality for training or testing force production capabilities at high velocities, notably for weaker populations.